1.

CLINICAL TRIALS PROCESS

1) Describe the clinical research process (phases, key roles,
key organizations).
2) Interpret a Statistical Analysis Plan.
3) Derive programming requirements from an SAP and an
annotated Case Report Form.

4) Describe regulatory requirements (principles of 21 CFR Part

11, International Conference on Harmonization, Good Clinical
Practices).

2.CLINICAL TRIALS DATA STRUCTURES

1) Identify the classes of clinical trials data (demographic,
lab, baseline, concomitant medication, etc.).
2) Identify key CDISC principals and terms.
3) Describe the structure and purpose of the CDISC SDTM data
model.

4) Describe the structure and purpose of the CDISC ADaM data

model.

5) Describe the contents and purpose of define.xml.

3.IMPORT AND EXPORT CLINICAL TRIALS

DATA
1) Apply regulatory requirements to exported SAS data sets
(SAS V5 requirements).

2) The FDA is defining standards for various types of electronic submissions to the
agency. A transport format developed by SAS Institute has been adopted as an FDA
standard for accepting and archiving data sets.

The Electronic Records; Electronic Signature Rule (21 CFR Part 11), or "the Esig
rule," is the US law that provides FDA the authority to identify types of submissions
that the agency will accept in electronic form. The Esig rule also states that persons
should consult the agency for details on how to proceed with electronic submissions.
Such details include method of transmission, media, file formats, and technical
protocols. The FDA issues guidance documents to reduce the need to consult for
details.

In January 1999, the FDA's Center for Drug Evaluation and Research (CDER) and
Center for Biologics Evaluation and Research (CBER) issued a joint guidance on
general considerations for electronic submissions, entitled Providing Regulatory
Submissions in Electronic Format - General Considerations . The guidance
describes the use of the SAS System XPORT format for electronic data sets.

The FDA plans to issue a series of guidance documents that will focus on specific
submission types. The first of this series was issued in January 1999, when CDER
issued a guidance on NDAs, entitled Providing Regulatory Submissions in Electronic
Format - NDAs .

Format of Electronic Data submission

You should provide data subsets in certain formats. Currently, we are able to accept
and archive datasets in SAS System XPORT transport format (Version 5 SAS
transport file). In circumstances when data are moved directly to a database or
special review tool, tagged ASCII file, specifically, standard generalized markup
language (SGML) and extensible markup language (XML), may be the appropriate
file format. At times, delimited ASCII files are also acceptable. See the individual
guidance for the specific submission type for the appropriate dataset format.

1. SAS System XPORT Transport Format (Version 5 SAS Transport Format)

SAS XPORT transport format, also called Version 5 SAS transport format, is an open
format published by the SAS Institute. The description of this SAS transport file
format is in the public domain. Data can be translated to and from this SAS transport
format to other commonly used formats without the use of programs from SAS
Institute or any specific vendor.
You should follow the recommendations in this section to create SAS transport files
that we can review and archive.
Version

In SAS, SAS XPORT transport files are created by PROC XCOPY in Version 5 of SAS
software and by the XPORT engine in Version 6 and higher of SAS Software. We are
unable to archive SAS Transport files processed by the CPORT engine.

You can find the record layout for SAS XPORT transport files in SAS technical support
TS-140. This document and additional information about the SAS Transport file
layout can be found on the SAS World Wide Web page athttp://www.sas.com/fda-
esub.

Transformation of Datasets

We use a variety of software tools to analyze the datasets. Stat/Transfer from Circle
Systems and DBMS/copy from Conceptual Software Inc. , are two programs used to
convert data to various formats used for analysis. SAS Viewer version 7 or higher is
used to open SAS transport files directly.

Naming SAS Transport Files

1. All SAS transport files should use .xpt as the file extension.
2. Compression of SAS Transport Files
SAS transport files should not be compressed. There should be one transport file
per dataset.

3. Content of Datasets and Organization

You should provide a single transport file for each dataset.

Many of the software tools used by the reviewers require datasets to be loaded into
random access memory (RAM) prior to opening the file. Therefore, dataset files
should be organized so that their size is generally less than 50 MB per file.

Datasets divided to meet the maximum size restrictions should contain the same
variable presentation so they can be easily merged, joined, and concatenated.

The datasets should be accompanied by data definition tables that include

metadata such as the variable name, a description of the variable, the type of
 variable (e.g. , number, character, date), and codes used in the dataset.

Variable names should be limited to 8 characters.

You should include a descriptive label up to 40 characters in the label header.

Further recommendations for content of SAS Transport files are provided in guidance
for each specific submission type.

4.MANAGE CLINICAL TRIALS DATA

1) Access DICTIONARY Tables using the SQL procedure.

Abstract
SAS users can quickly and conveniently obtain
useful information about their SAS session with a
number of readonly SAS data views called
DICTIONARY tables or DICTIONARY views. At any
time during a SAS session, information about
currently defined system options, libnames, table
names, column names and attributes, formats,
indexes, and more can be accessed and captured.
This paper will explore the purpose of DICTIONARY
tables and views, how they are accessed, and what
information is available to SAS users. Attendees
will learn how these important tables and views
can be applied using real-world code examples.

Introduction
The SAS System generates and maintains valuable
information at run time about SAS libraries, data
sets, catalogs, indexes, macros, system options,
titles, and views in a collection of read-only tables
called dictionary tables. Although called tables,
Dictionary tables are not real tables. Information is
automatically generated at runtime and the tables
contents are made available once a SAS session is
started. Dictionary tables and their contents
permit a SAS sessions activities to be easily
accessed and monitored. This becomes useful
when building software applications since the
information can be queried with the results acted
upon in a specific task such as in the allocation of
filerefs or librefs.

SAS users can quickly and conveniently obtain

useful information about their SAS session with a
number of read-only SAS data views called
DICTIONARY tables. At any time during a SAS
session, DICTIONARY tables can be used to capture
information related to currently defined libnames,
table names, column names and attributes,
formats, and much more. DICTIONARY tables are
accessed using the libref DICTIONARY in the FROM
clause of a PROC SQL SELECT statement. The name
of each DICTIONARY table and view along with its
purpose are presented below.

http://www2.sas.com/proceedings/sugi30/070-30.pdf

https://v8doc.sas.com/sashtml/proc/zsqldict.htm

http://www.lexjansen.com/pharmasug/2006/tutorials/tu03.pdf

http://www.sascommunity.org/seugi/SEUGI1996/KARLSSON1.pdf
2) Examine and explore clinical trials input data (find outliers,
missing vs. zero values, etc).

Missing values and outliers create difficulties for most analyses. Without
knowing the structure and details of missing data and outliers, problems may
arise when analysis is performed, such as undetected outliers which may bias
an analysis if these outliers are due to previously undetected errors.
Therefore, when receiving datasets with dozens (even hundreds) of variables
and thousands of observations, knowing the summary of missing values and
outliers can be helpful to identify any upstream problems that may have
crept into the data. Currently, a common method for this is to use PROC
MEANS, PROC UNIVARIATE, and PROC FREQ for each variable to detect the
missing values and outliers. If it is necessary, datasets are created to indicate
which records have missing values and/or outliers. Performing checks for
each variable, going through output for each procedure, and generating
datasets for each variable which have missing values and/or outliers is quite
time consuming, especially with large numbers of huge datasets. Individual
variable names need to be typed in for each dataset to find missing values
and/or outliers and output them to exception datasets. Additionally such
programs are not reusable due to different structures and variable names for
different studies. This can be very time-consuming.

5.TRANSFORM CLINICAL TRIALS DATA

1) Apply categorization and windowing techniques to clinical
trials data.

2) Transpose SAS data sets.

3) Apply 'observation carry forward' techniques to clinical
trials data (LOCF, BOCF, WOCF).

LOCF or Last Observation Carried Forward is a data imputation technique that not
surprisingly, carries the last assessed value forward to visits where a value is missing.
The LOCF method is often used for longitudinal data which is assessed over several
months or years. Despite all the effort to collect complete patient data for each single
time point these data eventually contain missing values. Missing values could for example
occur when a patient is lost to follow up or a patient does no show up for a single visit.

The LOCF method is also the technique which is mentioned in ICH E9 "Statistical
principles for clinical trials" where it says "[...] Imputation techniques, ranging from the
carrying forward of the last observation to the use of complex mathematical models, may
also be used in an attempt to compensate for missing data [...]"

There are multiple ways to program the LOCF method. Some of these methods are
explained in the following.
4) Calculate 'change from baseline' results.

5) Obtain counts of events in clinical trials.

http://analytics.ncsu.edu/sesug/2011/CC14.Go.pdf

6.APPLY STATISTICAL PROCEDURES FOR

CLINICAL TRIALS
1) Use SAS procedures to obtain descriptive statistics for
clinical trials data (FREQ, UNIVARIATE, MEANS, SUMMARY).

http://und.edu/org/rrvsug/presentations/

2) Use PROC FREQ to obtain p-values for categorical data (2x2

and NxP test for association).

3) Use PROC TTEST to obtain p-values for continuous data

(one-sample, paired and two-sample ttests).

4) Create output data sets from statistical procedures.

7.MACRO PROGRAMMING FOR CLINICAL

TRIALS
1) Create and use user-defined and automatic macro
variables.

2) Automate programs by defining and calling macros.

3) Use system options to debug macros and display values of

macro variables in the SAS log(MPRINT, SYMBOLGEN, MLOGIC,
MACROGEN).

8.REPORT CLINICAL TRIALS RESULTS

1) Use PROC REPORT to produce tables and listings for clinical
trials reports.

http://und.edu/org/rrvsug/presentations/ProcRpt_TimHunter.pdf

2) Use ODS and global statements to produce and augment

clinical trials reports .

http://support.sas.com/rnd/base/ods/odsdocument/

http://support.sas.com/resources/papers/proceedings12/273-2012.pdf

9.VALIDATE CLINICAL TRIAL DATA REPORTING

1) Explain the principles of programming validation in the
clinical trial industry.

http://www.sascommunity.org/wiki/Validation_of_SAS_Programs_in_an_FDA_Regul
ated_Environment

2) Utilize the log file to validate clinical trial data reporting.

http://www.lexjansen.com/pharmasug/2004/HandsOnWorkshops/HW07.pdf

http://www.sascommunity.org/mwiki/images/f/fb/Chang_and_soloff.pdf

http://www2.sas.com/proceedings/sugi26/p129-26.pdf

3) Use programming techniques to validate clinical trial data

reporting (PROC COMPARE,MSGLEVEL) .
 http://www.misug.org/uploads/8/1/9/1/8191072/churley_validation.pdf

4) Identify and Resolve data, syntax and logic errors.

http://www.lexjansen.com/scsug/2013/Strategies-and-Techniques-for-Debugging-
SAS-Program-Errors-and-Warnings-Kirk-Lafler.pdf